Viewpoint: Estimating the equity premium


  • The author is also affiliated with Arrowstreet Capital, LP, and NBER. This paper was presented in June 2007 as a State of the Art lecture at the Canadian Economics Association annual meeting at Dalhousie University in Halifax, Nova Scotia. A precursor was presented in January 2007 to the D-CAF Conference on Return Predictability at Copenhagen Business School. I am grateful to participants at both conferences, to John Cochrane, Jon Lewellen, Lubos Pastor, Ivo Welch, and Jeff Wurgler, and particularly to Angelo Melino for their thoughtful comments; to Bob Shiller, Moto Yogo, and my colleagues at Arrowstreet Capital, Sam Thompson and Tuomo Vuolteenaho, for joint research and many conversations on this subject; and to Alex Ogan, also of Arrowstreet Capital, for his able assistance with the data illustrated in figures 1 through 5. Email:


Abstract.  Finance theory restricts the time-series behaviour of valuation ratios and links the cross-section of stock prices to the level of the equity premium. This can be used to strengthen the evidence for predictability in stock returns. Steady-state valuation models are useful predictors of stock returns, given the persistence in valuation ratios. A steady-state approach suggests that the world geometric average equity premium fell considerably in the late twentieth century, rose modestly in the early years of the twenty-first century, and was almost 4% at the end of March 2007.


La théorie financière contraint le comportement diachronique des ratios de valorisation et relie transversalement les prix des actions au niveau de prime des actions sur les obligations. Voilà qui peut être utilisé pour renforcer la prédictibilité des rendements sur les actions. Les modèles de valorisation en régime permanent sont des prédicteurs utiles des rendements sur les actions, compte tenu du caractère stable des ratios de valorisation. Une approche en termes de régime permanent suggère que la moyenne géométrique mondiale de la prime des actions sur les obligations a chuté considérablement à la fin du 20e siècle, qu'elle a été modestement en hausse dans les premières années du 21e siècle, et qu'elle était à presque 4%à la fin de mars 2007.